The specific aim of this proposal is to determine whether In premenopausal and postmenopausal women blood levels of Insulin like Growth Factor-I, Insulin like Growth Factor Binding Protein -3 and Growth hormone are influenced by a low-fat high- carbohydrate diet. The general method to be employed will be to perform assays for growth factors and hormone on blood samples, and nutrient analysis of food records, collected from subjects taking part in our ongoing multicentre randomized dietary intervention trial. This trial is designed to test the hypothesis that intervention with a low-fat high-carbohydrate diet will, in women with extensive mammographic densities, over a 10 year period reduce the incidence of breast cancer by 29 percent. The trial is explanatory in that it seeks to determine if there is a biological effect of dietary fat reduction in terms of a reduction in breast cancer incidence. To meet this goal we have selected as participants highly motivated subjects who are at increased risk of breast cancer, we have provided them with a high level of assistance in making a dietary alteration, we follow them carefully to ensure the maintenance of dietary change and the correct identification of subjects who develop breast cancer, and we plan to analyse the results according to study group and dietary compliance. Recruitment of the 4615 subjects required for the trial was completed in November 1998. (A total of 4693 were randomized). Blood samples and food records from subjects enroled in the trial will be used to test the hypotheses given below about the effects of dietary intervention on growth factors and hormones associated with breast cancer risk. Modulation of these factors by dietary intervention would indicate potential mechanisms by which diet may influence risk of breast cancer. Although not the purpose of the research proposed here, the long term nature of the trial, the complete follow-up of all subjects, and the complete ascertainment of breast cancer, will ultimately allow any changes in blood markers found in the present research to be examined in relation to changes in breast cancer incidence.